Musk hit the face, Tesla to learn from the new domestic car manufacturing power?

Automatic driving, “grab 7” is about to start.

Musk hit the face, Tesla to learn from the new domestic car manufacturing power?

Automatic driving, “grab 7” is about to start.

The year 2021 is destined to be the first year of autonomous driving LIDAR.

But the debate about the technical divide of autonomous driving has never stopped, with some industry insiders describing autonomous driving as “as complicated as putting a man on the moon”. Tesla-led car companies are heading towards the ultimate route of vision solutions, while new domestic forces such as Azera, Xiaopeng and R Cars are firmly embracing LIDAR.

No conclusion has been reached on the technical merits of LIDAR and vision solutions, but the industry consensus is that a better solution is one that provides more and more accurate driving data.

As the car king Ford once said, if you can’t show a customer a better car, then all he wants is a faster horse. In terms of new car manufacturing, which is still in the horse racing stage, if you can’t give consumers a better one, they will go looking for another one. After all, there are really a lot of technology solutions to choose from.

Musk’s history of facepalming
In the matter of hot search, 50 million American microblogging netizens “Ma Yilong” will never admit defeat.

Remember when Tesla just entered China, Musk claimed that imported Tesla would never lower its price, later, domestic Tesla lowered its price while building a factory, and the owners of imported cars were honored to be the first crop of leeks, and Musk was hit in the face; later, Musk stood for dogcoin, and a group of believers in the cryptocurrency circle began to beat the drums like crazy, then Musk went on the Saturday Night Live show and admitted that dogcoin was a scam, and dogcoin fell from the top of the mountain. The dogcoin quickly fell from the top of the mountain, smashing the wallets of fans.

This time, it even lit the fire on his own car.

Tesla announced that “driver assistance features such as Autopilot on the North American best-selling Model 3 and Model Y will stop using radar sensors in favor of a camera-based system.” Musk then confirmed the news on Twitter and said that later Teslas made in China will also be eliminated.

Translation: we don’t even use radar anymore, you guys love to buy it.

Once the news came out, first the American people have voted with their feet.

Musk hit the face, Tesla to learn from the new domestic car manufacturing power?

The National Highway Traffic Safety Administration (NHTSA) website shows that Tesla Model 3 and Model Y models built on or after April 27 will no longer receive the agency’s safety test mark for many of their features.

Consumer Reports said it no longer lists the 2021 Tesla Model 3 as a “preferred model. According to Jake Fisher, senior director of the magazine, “It’s extremely rare for automakers to remove safety features from a car during the production process, even if only temporarily.”

And that result may soon be reversed again.

Recently, a Model Y engineering test car was photographed by passersby in Florida, USA, with an extra LIDAR on the roof rack.

The foreign media reporter further confirmed that the Model Y with California license plate is the official engineering test car of Tesla. The anonymous source further broke the news: “Tesla had an agreement with Luminar to purchase Luminar-built LIDAR sensors for engineering tests of autonomous driving technology.”

Musk hit the face, Tesla to learn from the new domestic car manufacturing power?

This may be Musk in the face of the history of a colorful stroke, because, Musk once on the LIDAR evaluation is: LIDAR are stupid x.

Musk in the matter of automatic assisted driving, is completely “visual party”. But this visual logic has not less let Tesla into the eye of the storm:.

May 7, 2016, Florida, a Tesla Model S in the Autopilot assisted driving state, hit a white van is turning across the road, Model S failed to identify the presence of the truck, because did not take braking measures, the driver died on the spot.

The post-accident investigation report showed that the camera identified the larger white van compartment as a “white cloud”, thus causing the collision. Similar accidents have occurred many times in the United States and at home.

Later, the white van was called “Tesla’s curse”, and the reason for this is that Tesla’s vision solution is defective in recognition.

Whether or not Tesla will use LIDAR in the future, there is actually a most critical issue behind this is that the new energy car companies, is to do addition, or subtraction.

The new domestic car-making force’s heavy bet: LIDAR
Musk once said that people are driving with their eyes, the vision program is acting as a function of the eyes.

But the problem is, there are differences between eyes.

Millimeter wave radar is characterized as being good at measuring lateral information, such as the distance and speed of objects ahead, but not good at measuring longitudinal information, such as the size and height of objects.

In the previous accident, Tesla’s millimeter wave radar had detected the presence of the truck car ahead, but it could not determine the height of the car, so the white car of the truck as a “white cloud” in the sky. The detection effect of millimeter wave radar is also related to the material of the object, in the fog, rain and snow and other bad weather and tunnel scenes is poor performance.

If the ultimate goal of autonomous driving is to replace human drivers, it can be divided into three core systems by function: perception (environmental awareness and positioning), decision making (intelligent planning and decision making), and execution (control execution).

To use an analogy, the perception layer is equivalent to the five human senses, seeing, hearing and sensing the surrounding environment, and feeding this information to the search and decision layer; and the decision layer is equivalent to the human brain, which has to process the information coming from the perception layer and then give instructions to the execution layer; then the execution layer is equivalent to the human limbs, which execute the instructions given by the brain.

Therefore, in the perception layer, it is necessary for the intelligent car to collect more and more complete information in order to assist the decision-making layer to make the right decision.

Autopilot, which relies purely on visual algorithms like Tesla, is not very suitable for roads with complex conditions like China. Not only is it unable to meet the redundancy and reliability requirements of intelligent driving, but it may also be unable to land under the increasingly strict regulatory constraints on intelligent driving.

Especially in a large number of Internet manufacturers into the node of the car, many people like to build a car and cell phone analogy, but the biggest difference between the two is that the phone card, dead still can restart, but once the car on the road problems, there may never be a restart opportunity. Therefore, a number of new domestic forces standing opposite Tesla are using LIDAR in their assisted driving programs.

The ET7 of Azure add LIDAR, and the Extreme Fox in cooperation with Huawei also use radar, traditional car manufacturers, such as Audi, BMW, Nissan, Toyota, etc. have also announced that they will apply LIDAR, but have not yet announced a specific timetable; as the rise of new forces in car-making suppliers Huawei is insisting on the use of LIDAR, and boasts that the cost of 96-line LIDAR will be reduced to less than $200, before the same level The price of LIDAR is more than $1000.

Among them, R Auto has already used the Luminar LIDAR of Tesla’s road test. With the empowerment of R-TECH’s high energy wisdom body, R Auto’s new model ES33 uses the industry’s top Luminar LIDAR solution and becomes the world’s first mass production model.

Musk hit the face, Tesla to learn from the new domestic car manufacturing power?

Compared with conventional LIDAR, Luminar LIDAR can generate the most fine three-dimensional point cloud information, using 1550NM light source in line with the safety of the human eye, the detection distance can reach 500m, in the rain, snow and fog and other bad weather still have good performance, can see farther, clearer, three-dimensional description of the obstacles more accurate, as if the car has a brighter pair of eyes.

This is precisely the opposite of Tesla’s choice, Tesla is constantly subtracting: significantly compress the manufacturing process, reduce the test link, etc., all in order to reduce costs. If even the millimeter wave radar is removed, then it is likely to encounter the market vote with their feet, after Tesla’s “brake door” has been proven by the market, the incident of safety consumers absolutely zero tolerance, Tesla April Model 3 sales of 6264 units, and March 25,327 units, compared to the ring Tesla Model 3 sales in April 6264 units, compared with March 25,327 units, a year-on-year plunge of nearly 75%, and may continue to decline.

So, in order to make the whole car system become safe and reliable, enhance the redundancy of the perception system has become the solution of many car companies, that is, to do “addition”.

Because, to give an extra layer of security, at no time is redundant.

What should a real intelligent tram look like?
How smart should you expect the tram to be?

In the morning and evening commute, it can help you identify the various road conditions in the congested morning and evening rush, as well as the pedestrians who do not follow the traffic rules; or, it can help you not to drive fatigued due to visual burnout on the highway without changes on holidays, but also to arrive safely and smoothly in the distance, or, in the case of traffic and weather is not so high quality, still do not need you to use too much energy a forever. A hard-working and smart machine.

According to the research report of the E-Car Research Institute, in recent years, Chinese car-buying users have continued to increase their attention to intelligence and technology, and in the first three quarters of 2020, intelligent technology ranked third in terms of car-buying concern factors for men and second in terms of car-buying concerns for women.

Smart, technology, is a new wave of consumer concern common to all consumers. And which technology tree to light up becomes the key.

Backed by SAIC Group, the high-end electric car brand R Auto is planning the layout in this way.

R-TECH of R-Car has created a “six-fold fusion sensory system” for smart cars, including LIDAR, 4D imaging radar, vision camera, ultrasonic radar, high-precision map, 5G V2X, with all-weather, multi-dimensional and super-visual perception capabilities, which can be called the “six-vein sword” in the field of smart driving. The “Six Veins of Divine Sword”.

But the so-called car intelligence is by no means a simple pile of material game, how to achieve the optimal effect of each ring, so that each module coupling is the key to intelligent change, or the difficult point.

R car advanced intelligent driving solution PP-CEMTM, can combine Pixel + Point Cloud image point cloud perception fusion system, can more accurately predict the behavior and trajectory of pedestrians, vehicles and other obstacles. r car will also be the first to mass production application of nvidia Orin super computing chip, with the full stack of self-developed super environment model algorithm, to achieve the perception of the fusion, will The safety of smart driving is raised to a whole new level.

As a result, it won’t confuse the white cargo box of the truck in front of it with the sky, and it won’t be helpless in dark and narrow tunnels, and it will be able to recognize every dangerous scene under the overpass. No matter in the “Chinese” complex traffic scenario, or encounter rain, snow, fog and other extreme weather scenarios, users can be at ease, calmly deal with.

In addition to safety and intelligence, we also need to return to the more primitive tram mileage anxiety. In order to alleviate such anxiety, R Auto has also designed an intelligent charging management solution that can extend battery life and improve charging safety. At the same time, it combines vehicle data, charging pile information, road information and self-selected preferences to automatically plan out charging suggestions for long-distance routes, such as charging times, charging times and charging routes.

In addition, as a 5G car, through the high-speed and real-time network communication brought by the 5G network, R Auto’s SuperMap will become as convenient and quick to find charging piles as finding a restaurant through VWAP after accessing 250,000 charging piles including those containing State Grid, Southern Grid, Anyue Charging, Tesco, and Star Charging. What’s more, you can see people’s real-time evaluation of the charging piles and the surrounding leisure and entertainment information. Charging will no longer be the most boring step in daily travel.

When the once-in-a-century wave of change in the automotive industry sweeps in, for the car companies that have been on the road, win to the end, should be the safety, intelligence and humane triple control, only have a certain ability of the car companies I’m afraid it is difficult to go steady and far.

At present, the new domestic forces have been actively responding to the battle.

Posted by:CoinYuppie,Reprinted with attribution to:
Coinyuppie is an open information publishing platform, all information provided is not related to the views and positions of coinyuppie, and does not constitute any investment and financial advice. Users are expected to carefully screen and prevent risks.

Like (0)
Donate Buy me a coffee Buy me a coffee
Previous 2021-05-30 22:12
Next 2021-05-30 22:25

Related articles